Commutators of PAR-1 signaling in cancer cell invasion reveal an essential role of the Rho–Rho kinase axis and tumor microenvironment

Commutators of PAR-1 signaling in cancer cell invasion reveal an essential

role of the Rho–Rho kinase axis and tumor microenvironment

Quang-De Nguyen1,6, Olivier De Wever1,6, Erik Bruyneel2, An Hendrix2, Wan-Zhuo Xie3,4, AlainLombet5, Martin Leibl3, Marc Mareel2, Frank Gieseler3, Marc Bracke2 and Christian Gespach*,1

1INSERM U673, Molecular and Clinical Oncology of Human Solid Tumors, Hopital Saint-Antoine, 184 Rue du faubourg Saint-Antoine, 75571 Paris Cedex 12, France; 2The Laboratory of Experimental Cancerology, Ghent University Hospital, B-9000 Ghent,Belgium; 3Department of Internal Medicine, University of Kiel, Schittenhelmstr. 12, 24105 Kiel, Germany; 4First Affiliated Hospital,College of Medicine, Zhejiang University, Zhejiang, China; 5CNRS UMR8078, Hopital Marie Lannelongue, 92350 Le Plessis-Robinson, France

We recently reported that proteinase-activated receptorstype I (PAR-1) are coupled to both negative and positiveinvasion pathways in colonic and kidney cancer cellscultured on collagen type I gels. Here, we found thattreatments with the cell-permeant analog 8-Br-cGMP andthe soluble guanylate cyclase activator BAY41-2272, andRho kinase (ROK) inhibition by Y27632 or a dominantnegative form of ROK lead to PAR-1-mediated invasionthrough differential Rac1 and Cdc42 signaling. Hypoxiaor the counteradhesive matricellular protein SPARC/BM-40 (SPARC: secreted protein acidic rich in cysteine)overexpressed during cancer progression also commutatedPAR-1 to cellular invasion through the cGMP/proteinkinase G (PKG) cascade, RhoA inactivation, and Rac1-dependent or -independent signaling. Cultured primarycancer cells isolated from peritoneal and pleural effusionsfrom patients with colon cancer or other malignant tumorsharbored PAR-1, as shown by RT–PCR and FACSanalyses. These malignant effusions also contained highlevels of activated thrombin and fibrin, and induced aproinvasive response in HCT8/S11 human colorectalcancer cells. Our data underline the essential role of thetumor microenvironment and of several commutatorstargeting cGMP/PKG signaling and the RhoA–ROKaxis in the control of PAR-1 proinvasive activity andmetastatic potential of cancer cells in distant organs andperitoneal or pleural cavities. We also add new insightsinto the mechanisms linking the coagulation mediatorsthrombin and PAR-1 in the context of blood coagulationdisorders and venous thrombosis often observed in cancerpatients, as described in 1865 by Armand Trousseau.Oncogene (2005) 24, 8240–8251. doi:10.1038/sj.onc.1208990;published online 8 August 2005

Keywords: PKG; hypoxia; SPARC/BM-40; tenascin-C;peritoneal and pleural effusions; metastasis and throm-bosis

Introduction

The multifunctional serine protease thrombin exertsdirect actions on a wide variety of cell types, includingepithelial cells, endothelial cells, vascular smooth musclecells, leukocytes, neurons, and glial cells (Coughlin,2000). Thrombin is a potent mitogen for fibroblasts andepithelial cells, and potentiates the proliferative re-sponses of tumor cells to classical growth factors, suchas epidermal growth factor and insulin (Van Obber-ghen-Schilling et al., 1995). Both thrombin and protei-nase-activated receptors type I (PAR-1) are expressed ininvasive cancer cell lines and breast carcinoma biopsyspecimens (Wojtukiewicz et al., 1995; Even-Ram et al.,1998, 2001; Nguyen et al., 2002; Darmoul et al., 2003).However, the signaling pathways triggered by thrombinand PAR-1 signaling during tumor progression andinvasion remain poorly documented.Thrombin interacts with the G-protein-coupled pro-

tease-activated receptors PAR-1, -3, and -4. ActivatedPAR-1 are coupled via several members of the hetero-trimeric G-proteins Gao/i, Ga12/13, and Gaq totransduce a substantial network of signaling pathways(Coughlin, 2000). We have recently demonstrated thatPAR-1 are functionally connected to both negative andpositive invasion pathways in colon and kidney cancercells (Faivre et al., 2001; Nguyen et al., 2002). In thecollagen type I substratum, PAR-1 and the pertussistoxin (PTx)-sensitive Gao/i subunits were shown toexert a dominant invasion suppressor role towardseveral proinvasive pathways controlled by oncogenesand tumor-secreted growth factors (Faivre et al., 2001).Similar findings were reported by Kamath et al. (2001)for the inhibition of invasion and migration in the highlyinvasive MDA-MB231 breast cancer cell line, throughPAR-1- and Gai-dependent pathways. Conversely, wehave shown that neutralization of Gao/i signaling byPTx led to the proinvasive activity of thrombin andPAR-1 through the Ga12/13/RhoA cascade, myosinlight chain (MLC) phosphorylation, and activation ofthe actomyosin system (Kureishi et al., 1997; Nguyenet al., 2002). We presented evidence that the proinvasivepotential of PAR-1 in collagen type I is also revealed by

Received 13 April 2005; revised 15 June 2005; accepted 1 July 2005;published online 8 August 2005

*Correspondence: C Gespach; E-mail: [email protected] two authors contributed equally to this work

Oncogene (2005) 24, 8240–8251& 2005 Nature Publishing Group All rights reserved 0950-9232/05 $30.00

www.nature.com/onc

inhibition of RhoA GTPase by RhoD, C3 exoenzyme,the dominant negative N19-RhoA, and cGMP-elevatingagents sodium nitroprusside and guanylin, actingthrough soluble and membrane-bound guanylate cy-clases, respectively (Nguyen et al., 2002). In this case,PAR-1 is connected with the Gaq/phospholipaseCb–Ca2þ /calmodulin-MLC kinase (CaM-MLCK) cas-cades that bypass RhoA blockade (Nguyen et al., 2002).Thus, we defined a new function for the small GTPasesRhoA and RhoD acting as a molecular switch control-ling the negative and positive invasive pathwaystriggered by PAR-1 through PTx-independent hetero-trimeric G-proteins. Our data were reminiscent of thedemonstration that PAR-1 induce invasion of Matrigelby breast cancer cells through requirement of avb5integrins (Even-Ram et al., 1998, 2001), suggesting thatthe PAR-1 invasive potential is controlled by thematricellular context (Nguyen et al., 2002).Important features of invasive cancers include dra-

matic changes in the tumor microenvironment andinfiltration by immune and mesenchymal cells andmyofibroblasts, and remodeling of several extracellularmatrix (ECM) components such as tenascin-C (TN-C)and SPARC/BM-40 (SPARC: secreted protein acidicrich in cysteine; Porte et al., 1995; Ledda et al., 1997;Thomas et al., 2000). In addition, hypoxia and increasedproduction of the free radical nitric oxide (NO) play acentral role in invasive tumor growth, angiogenesis, andmetastasis. Inducible isoforms of nitric oxide synthase(iNOS) are overexpressed in human colonic tumors,while genetic ablation or inhibition of iNOS reducesaberrant crypt foci, colon tumor formation, and lungtumorigenesis (Ahn and Ohshima, 2001 ; Kisley et al.,2002). Here, we examined the impact of the RhoAeffector Rho kinase (ROK), the soluble guanylatecyclase (sGC) activator BAY41-2272 (Stasch et al.,2001), as well as hypoxia and matricellular proteinsSPARC and TN-C, on the commutation of negative andpositive signaling pathways controlled by PAR-1 incancer cell invasion. We demonstrate that these mole-cular and pharmacological interventions targeting thecGMP and RhoA–ROK signaling pathways, as well asother key parameters linked to the tumor microenviron-ment and metastatic process, exert a permissive role onPAR-1 proinvasive activity.

Results

Promotion of PAR-1 proinvasive activity by inhibitionof the RhoA–ROK axis

The Rho-GTPase subfamily members are involved inthe regulation of cell shape, cytoskeletal reorganization,and motility through activation of the actin microfila-ment network and establishment of intercellular andadhesive contacts with ECM components. RhoAregulates the formation of stress fibers, focal adhesions,and cell contraction, Rac1 induces the formation oflamellipodia and membrane ruffling, whereas Cdc42 isinvolved in filopodia formation (Ridley, 2001). We have

shown that small GTPases RhoA and RhoD act asmolecular switches at the negative and positive invasionpathways controlled by PAR-1 in kidney and coloncancer cells (Nguyen et al., 2002). Moreover, the RhoAantagonist RhoD and the RhoGTPase inhibitor C3exoenzyme promoted the PAR-1 proinvasive activitythrough a signaling cascade using Gaq, phospholipase-Cb, and CaM-MLCK. In this signaling network,cGMP-elevating agents (sodium nitroprussiate, guany-lin) and protein kinase G (PKG) act as RhoA inhibitors,as previously reported (Sauzeau et al., 2000; Sawadaet al., 2001).In order to determine the molecular specificity of the

PAR-1 commutation by the Rho-GTPase family mem-bers, we have established a series of human HCT8/S11colon cancer cell lines expressing dominant negativemutants of RhoA, Rac1, and Cdc42, as well as theconstitutively activated form of RhoA. As previouslydemonstrated in kidney cancer cells, parental and stablytransfected HCT8/S11 cells by DN-Rho GTPase andDC-RhoA vectors express PAR-1 receptors as a 90 kDaband downregulated by specific PAR-1 antisense (top ofFigure 1a; Nguyen et al., 2002). Similarly, invasion oftype I collagen gels was induced by the PAR-1 agonistTRAP in HCT8/S11 cells stably transfected with theinterfering mutant DN-RhoA (T19N), as shown inFigure 1, and previously in kidney epithelial cellsMDCK-T23 (Nguyen et al., 2002). Such a commutationcannot be accomplished by the constitutively active DC-RhoA (G14V) or dominant interfering mutants DN-Rac1 (T17N) and DN-Cdc42 (T17N) in colorectalcancer cells HCT8/S11 (Figure 1a), as well as in kidneyepithelial cells MDCK-T23 expressing interfering mu-tants DN-RhoA and DN-Rac1 (Jou and Nelson, 1998)under the tetracycline-repressible transactivator (notshown).Since RhoA and its direct effector ROK regulate the

dynamic reorganization of the cytoskeletal actin net-work leading to activation of the actomyosin system,stress fibers assembly, and cell motility (Amano et al.,1997), we decided to examine the role of the RhoA–ROK axis in the regulation of the thrombin PAR-1receptor proinvasive behavior. Here, we found that theROK inhibitor Y27632 reveals the proinvasive activityof PAR-1 through Rac1-dependent and Cdc42-indepen-dent signaling pathways (Figure 1b and c). In order toverify this observation, we next established stablytransfected cell lines HCT8/S11-DN-ROK and -DC-ROK expressing dominant negative and constitutivelyactivated forms of ROK, respectively (Amano et al.,1997). As shown in Figure 1d, both parental and DN-ROK-transfected HCT8/S11 cells were noninvasiveunder control conditions. Interestingly, the dominantnegative form of ROK established the proinvasiveactivity of PAR-1, but failed to produce the sameresponse following activation of the G-protein-coupledPAF-R. As expected (Rodrigues et al., 2001), DN-ROKabrogated the proinvasive activities of pS2 and leptin(Figure 1d), whereas DC-ROK was ineffective. Weconfirmed that this dominant active mutant wasfunctionally competent in our system since DC-ROK

Hypoxia, SPARC, Rho/ROK, and PAR-1 proinvasive activityQ-D Nguyen et al

8241

Oncogene

abrogated insulin-induced invasion. Indeed, activationof the RhoA–ROK axis was shown to negate insulinsignaling via association of insulin receptor substrateIRS-1 with ROK (Begum et al., 2002). The half-maximal effect of insulin on collagen invasion wasobserved at the potency EC50 of 0.6 nM consistent withactivation of insulin receptors in intestinal epithelialcells. Alternatively, insulin may act through additionaltargets by modulating positively cGMP generation andPI3-kinase pathways (Emami and Perry, 1984; Begum

et al., 2002) involved in cancer cell invasion (Kotelevetset al., 1998). Interestingly, PAR-1 commutation inducedby DN-RhoA, ROK inhibition, and DN-ROK (Figure1a, b, and d) is not associated with significant changes inPAR-1 expression in HCT8/S11 cells incubated withTRAP and under control conditions (Figure 1a and b,top panels).Thrombin is an endogenous PAR-1, PAR-3, and

PAR-4 agonist. PAR-2 is activated by trypsin, tryptase,and coagulation factors VIIa and Xa, but not bythrombin. There is evidence that PAR-2 is involved incolon cancer cell growth through transactivation ofEGF-R receptors (Darmoul et al., 2003). As previouslyobserved in kidney cancer cells (Nguyen et al., 2002),PAR-2 and PAR-4 agonists are ineffective in inducingcollagen type I invasion in HCT8/S11 cells incubatedwith the Rho-GTPase inhibitor C3T and under controlconditions (not shown). In contrast, the PAR-1 com-mutators C3T and Y27632 were effective in determiningthe invasive phenotype in human colorectal cancer cellsHCT8/S11, HT29, and HCT116 (not shown). More-over, both PAR-2 and -4 activating peptides failed todownregulate invasiveness determined by the trefoilpeptide pS2 (not shown). Taken together, our dataconfirm that invasive growth of human colon cancercells in collagen type I is not controlled by PAR-2 and -4agonists (Nguyen et al., 2002).

Promotion of PAR-1 proinvasive activity by directactivation of soluble guanylate cyclase

The importance of the cGMP pathways in RhoAinactivation is well established (Sauzeau et al., 2000;Sawada et al., 2001), while its significance in theneoplastic progression and cancer cell invasion is stillobscure. We have shown that PAR-1-mediated invasivepotential of human colorectal cancer cells in collagentype I gels can be achieved with the cGMP-elevatingagents sodium nitroprusside and guanylin acting asPAR-1 commutators through soluble and membrane-bound guanylate cyclases, respectively (Nguyen et al.,

Figure 1 Induction of PAR-1 proinvasive activity by inhibition ofthe RhoA–ROK axis. (a) Invasion of collagen type I gels by HCT8/S11 human colorectal cancer cells stably transfected by expressionvectors encoding dominant negative forms of small GTPases (DN-RhoA, DN-Rac1, and DN-Cdc42), a constitutively activated formof RhoA (DC-RhoA), or the control pcDNA3 vector (controlsham-transfected cells). Cells were incubated at 371C for 24 h in thepresence or absence of 10mM PAR-1 agonist TRAP. (b) HCT8/S11cells were treated with 10mM TRAP and ROK inhibitor Y27632(10 mM), either alone or combined. Western blot analysis (a and b,top panels) showed the expression of PAR-1 (90 kDa) in HCT8/S11cells submitted to the same experimental conditions. (c) HCT8S/11cells expressing the DN-Rac1 and DN-Cdc42 mutants were treatedwith TRAP and Y27632 (each at 10 mM), either alone or combined.(d) Parental HCT8/S11 cells (Cont) and their counterparts stablytransfected with either DN-ROK (lacking the RB/PH domain) orDC-ROK (containing exclusively the catalytic CAT domain) weretreated with the proinvasive agents TRAP (10mM), pS2 (100 nM),leptin (100 ng/ml), and insulin (100 nM). PAR-1 and PAF-Rresponses were compared using 0.1 mM PAF


8242

Oncogene

2002). Here, we report that the direct activator of sGCBAY41 promotes a permissive signal for the proinvasiveactivity of PAR-1 (Figure 2a). The sGC activatorBAY41 at 10 nM–1 mM induced the invasive capacity ofHCT8/S11 cells exposed to a fixed concentration ofTRAP (10 mM). Similarly, the PAR-1 agonist TRAPdose-dependently (10 nM–10 mM) stimulated collagentype I invasion by HCT8/S11 cells treated with a fixedconcentration of BAY41 (1 mM), but was ineffectivewhen tested alone at the same concentrations (notshown). The invasive response induced by TRAP in thepresence of BAY41 was blocked by PLCb and CaM-MLCK inhibitors (U73122 and KT5926), but wasinsensitive to the ROK inhibitor Y27632, as expected(Figure 2b). Thus, the two PAR-1 commutators BAY-

41 and ROK inhibitor Y27632 are not mutuallyexclusive and instead produce, as expected, additiveresponses to PAR-1 activation. In the presence ofBAY41, PAR-1 operate through PLCb/CaM-MLCK-dependent proinvasive pathways. As control, wechecked that BAY41 dose-dependently (0.1 and 1 mM)abrogated cellular invasion induced by leptin and thetrefoil factors pS2 and ITF (not shown), both actingthrough Rho-dependent signaling pathways (Rodrigueset al., 2001). As shown in Figure 2c, the proinvasiveactivity determined by PAR-1 and the sGC activatorBAY41 was blocked by the DN-Rac1 interferingmutant, but was insensitive to DN-Cdc42.

Promotion of PAR-1 proinvasive activity by thematricellular proteins SPARC and TN-C

Interactions between cancer cells, tumor stromal cells,and matricellular molecules participate in the mechan-isms of invasive growth in malignant tumors. Invasivecell migration is controlled by a vast array ofextracellular factors, which require the coordinatedactivity of cell adhesion molecules facing ECM compo-nents. Adhesion molecules and cell–matrix communica-tions control immediate cellular responses, such asactivation of the cytoskeleton, cell shape, and motility,as well as delayed responses, including gene transcrip-tion. SPARC or BM-40, a secreted protein, acidic andrich in cysteine, also known as osteonectin, is a 43-kDaECM glycoprotein involved in cell–ECM interactionsduring wound healing, tissue remodeling, and cancerprogression (Bradshaw and Sage, 2001). Both SPARCand TN-C are ECM components highly expressed incancers of the gastrointestinal tract and a wide range ofhuman malignant neoplasms. Recently, TN-C wasshown to downregulate activated forms of GTP-boundRhoA in fibroblasts and HCT8/S11 epithelial cells (DeWever et al., 2004), suggesting that this matricellularcomponent can function as a PAR-1 signaling commu-tator (Nguyen et al., 2002). We have therefore investi-gated the proinvasive potential of PAR-1 in HCT8/S11cells cultured on collagen type I in association with theseECM components. As expected, both SPARC and TN-C-containing collagen type I matrices promoted thePAR-1 invasive potential (Figure 3a), as demonstratedfor TN-C and HGF, another proinvasive agent actingthrough the Met oncogene (De Wever et al., 2004).Preincubation of HCT8/S11 cells plated on SPARC for15min with the PKG inhibitor KT5823 (2 and 20 mM)abrogated invasion promoted by activated PAR-1 (n¼ 3experiments, data not shown). The matricellular proteinSPARC induced a dramatic neutralization of RhoAactivity in HCT8/S11 cells plated on top of SPARC-containing matrix, as compared to collagen type I(Figure 3b). Conversely, HCT8/S11 cells cultured onSPARC exhibit a twofold increase in Rac1 activity(from 1 to 2.05 relative intensity), which was not furtherelevated in the presence of the PAR-1 agonist TRAP(1.96 relative intensity). As shown in Figure 4a, this wasassociated with decreased cellular adhesion and multiplecellular extensions in HCT8S11 cells cultured on

Figure 2 Induction of PAR-1 invasive activity by the sGCactivator BAY41. (a) Induction of cellular invasion by TRAP(10mM, filled circles) in HCT8/S11 cells exposed for 24 h to variousconcentrations of BAY41 (1 nM–1 mM, open triangles). (b) Effectsof inhibitors targeting ROK (Y27632, 10 mM), PLCb (U73122,1mM), and Ca-MLCK (KT5926, 20 nM) on invasion induced by10mM TRAP combined with 1mM BAY41 (Control: Cont). (c)HCT8/S11 cells stably expressing the DN-Rac1 and DN-Cdc42were treated with TRAP (10mM) combined with BAY41 (1mM)


8243

Oncogene

SPARC, as compared to control collagen type I (whitearrowheads), and 82% reduction of focal adhesions(Po0.05), as monitored by vinculin staining (whitearrows in control cells). These results are consistent withprevious data showing that SPARC exerts counter-adhesive effects through the disassembly of focaladhesion plaques in endothelial cells (Murphy-Ullrichet al., 1995). Most interestingly, we observe in Figure 4bthat TRAP-stimulated HCT8/S11 cells enteringSPARC-containing collagen type I gels are not blockedby dominant interfering Rac1 and Cdc42 mutants. Inthis situation, we hypothesize that PAR-1 proinvasiveactivity is mainly driven by the RhoA status since RhoAinactivation can be achieved through serine phosphory-lation and translocation from the plasma membrane tothe cytosol. Such an event is induced by the cGMP-dependent protein kinase PKG (Sauzeau et al., 2000;Sawada et al., 2001).

Figure 3 Induction of PAR-1 proinvasive activity by the matricel-lular proteins SPARC and TN-C. (a) Invasive potential of HCT8/S11 cells cultured on collagen type I alone (Coll I), or a mixture ofcollagen type I containing either SPARC (10 nM) or TN-C (50 pg/ml). (b) RhoA and Rac1 activity in cell lysates prepared fromHCT8/S11 cells grown in serum-free conditions and seeded onsubstrates coated with collagen type I (Coll I) or SPARC (10 nM), inthe presence or absence of TRAP (10mM). Activated GTP-boundRhoA and Rac1 were assayed using agarose beads linked to theRho-binding domain of Rhotekin or the Rac effector p21-activatedkinase PAK-1. Relative intensity of the RhoA-GTP and Rac1-GTPimmunoreactive bands was quantified, according to total RhoA andRac1 levels detected by direct Western blots (arbitrary units, AU),using the ImageQuant Software (Amersham Biosciences). Data arerepresentative of 1–4 other experiments

Figure 4 Induction of a RhoA neutralization phenotype, activa-tion of the actin cytoskeleton, and focal adhesion disassembly bythe matricellular protein SPARC. (a) HCT8/S11 cells were seededon collagen type I (10mg/ml) or SPARC (10 nM). After 24 h,adhesive cells were examined by phase-contrast microscopy(bar¼ 40mm). Cells adherent to collagen type I showed typicalcellular spreading with formation of focal adhesions and stressfibers. SPARC remarkably induced actin belts and formation ofcellular extensions (white arrowheads), as evidenced by F-actinstaining (white arrowheads) and reduced focal adhesions by 82%(Po0.05*) assayed by vinculin staining (white arrows,bar¼ 20 mm). The number of focal adhesions was measured by adirect counting on the computer screen. Data are means7s.e.m.from at least 10 different fields. (b) HCT8/S11 cells stablyexpressing DN-Rac1 and DN-Cdc42 were cultured on top ofcollagen type I gels containing 10 nM SPARC and treated by TRAP(10mM)


8244

Oncogene

Promotion of PAR-1 proinvasive activity by hypoxia

Low oxygen availability during intratumoral hypoxia isa general feature of growing tumors. Hypoxic environ-ment has been shown to play a major role in tumorangiogenesis, cancer growth, and metastasis. Thehypoxia-dependent angiogenic switch in human solidtumors is an early event linked to the secretion of highlevels of VEGF and several angiogenic factors by cancercells, as well as by endothelial cells or other cell types inthe tumor stroma (Andre et al., 2000). The NO/cGMP/PKG and calcium-dependent signaling pathways arealso involved in the regulation of tumor cell invasivenessand angiogenesis linked to endothelial cell proliferation,morphogenesis, and tumor invasion (Kawasaki et al.,2003). These observations led to the hypothesis thathypoxic conditions could exert a direct or a permissiverole on the PAR-1 invasive potential. As shown inFigure 5a, low oxygen levels were inefficient in inducinga spontaneous invasive phenotype. Under hypoxia, thePAR-1 agonist dose-dependently promoted this re-sponse, according to the potency EC50¼ 0.1 mM TRAP.Hypoxic conditions were associated with significant 2.2-and 2.5-fold increase in cGMP levels in HCT8/S11 cellscultured under control conditions (10.6 versus23.6 pmol/106 cells, Po0.05) or in the presence of thephosphodiesterase inhibitor IBMX (14 versus 35.6 pmol/106 cells, Po0.05), as shown in Figure 5b. This hypoxia-induced cGMP generation was not further elevated inthe presence of TRAP (not shown). Hypoxia-mediated

PAR-1 activity was abrogated by inhibitors of the Gaq–PLCb–CaM-MLCK cascade (U73122 and KT5926),but was unaffected by inhibitors of the RhoA–ROKaxis, as expected (Figure 6a). Hypoxia induced aremarkable downregulation of RhoA activity in thepresence or absence of TRAP, a response associatedwith increased Rac1 activation (1.6-fold) in the presenceof the PAR-1 agonist (Figure 6b). In agreement,invasiveness induced by PAR-1 activation in hypoxicHCT8/S11 cells was not reversed by DN-RhoA andDN-Cdc42, but was impaired by the interfering form ofRac1 (DN-Rac1). Our data are reminiscent of theobservation that hypoxia sensitizes cancer cells to HGFproinvasive activity (Pennacchietti et al., 2003).Since both hypoxia and TN-C/SPARC expression

correlate with tumor progression, we next examined theimpact of oxygen deficiency on the expression of thesetwo matricellular proteins. Our data in Figure 7a and bprovide the first evidence that the SPARC and TN-Cgenes are expressed in human colorectal cancer cells atboth transcriptional and protein levels. While both TN-C and SPARC transcripts were significantly elevatedunder hypoxia (5.8- and 6.2-fold induction, respectively,Po0.05), their protein levels were found downregulated10-fold (Po0.05) in cellular extracts prepared fromhypoxic HCT8/S11 cells. Since both TN-C and SPARCproteins were undetectable in the conditioned media andextracellular matrices prepared from normoxic andhypoxic cells (not shown), one can postulate that thepermissive action of hypoxia on the PAR-1 proinvasiveaction is mediated through cGMP-dependent and TN-C/SPARC-independent mechanisms. In line with this,recent studies indicate that both Ras activation andTGFb are implicated in TN-C secretion and matrixdeposition (Maschler et al., 2004). As control, weconfirmed that hypoxic conditions were associated witha remarkable 7.8-fold induction of the hypoxia-induci-ble factor 1a (HIF-1a) subunit (Figure 7b). To ourknowledge, there is no information on TN-C andSPARC degradation by the proteasome. Interestingly,we found that the proteasome inhibitor MG132(100 nM, 2 h preincubation of HCT8/S11 cells beforethe 24 h invasion assay in the presence of MG132)selectively obliterates cellular invasion determined byleptin or PAR-1 activation in the presence of BAY41(not shown).

Implication of the cGMP–PKG pathways in PAR-1proinvasive activity

Since hypoxic conditions promote the proinvasiveactivity of PAR-1 and significant elevation of cGMPlevels in HCT8/S11 cells (Figure 5), we next examinedthe possible implication of PKG in this commutation.As observed above for the PAR-1 commutator SPARC,15min preincubation of HCT8/S11 cells with the PKGinhibitor KT5823 (2 and 20 mM) abrogated the PAR-1proinvasive activity determined by subsequent treatmentwith TRAP in hypoxia (Figure 8a). This blockade wasnot observed when KT5823 and TRAP were addedsimultaneously in hypoxic conditions (not shown).

Figure 5 Induction of PAR-1 invasive activity and cGMPgeneration by hypoxia. (a) HCT8/S11 cells were cultured inhypoxic conditions for 24 h in the presence or absence of increasingconcentrations of the PAR-1 agonist TRAP (1 nM–30mM). Inva-sion assays were performed on type I collagen gels. (b) HCT8/S11cells were cultured in normoxia or hypoxia on collagen type I gelsfor 24 h, in the presence or absence of the phosphodiesteraseinhibitor IBMX (1mM). cGMP generation was quantified byradioimmunoassay


8245

Oncogene

Notably, cellular invasion induced by PAR-1 activationin the presence of the sGC/cGMP activator BAY41 innormoxia was blocked by simultaneous addition ofKT5823 (2 and 20 mM), as shown in Figure 8b. Tofurther demonstrate the direct implication of the cGMP/PKG cascade in the PAR-1 proinvasive activity, weconfirmed that 8-Br-cGMP (5 mM) behaved as a selectivePAR-1 commutator (Figure 8c): (1) this cell-permeantcGMP analog was ineffective in promoting the pro-invasive activity of PAF and Met receptors (Kotelevetset al., 1998; De Wever et al., 2004) and (2) the invasiveresponse determined by 8-Br-cGMP and PAR-1 activa-tion was blocked by 15min preincubation of HCT8/S11cells with the PKG inhibitor (data not shown, n¼ 3experiments).

Figure 6 Induction of PAR-1 proinvasive activity by hypoxia ismediated by the Gaq/PLCb/Ca-MLCK cascade and small GTPaseRac1. (a) The invasive phenotype of HCT8/S11 cells in collagengels was induced by hypoxic conditions. Cells were incubated for24 h in control conditions (Cont: no addition) or in the presence ofTRAP (10mM), either alone or combined with selective inhibitorsof the RhoA/ROK cascade (C3T, 5 mg/ml; Y27632, 10 mM), PLCb(U73122, 1mM), and Ca-MLCK (KT5926, 20 nM). (b) Activity ofRhoA and Rac1 in HCT8/S11 cells cultured under normoxic andhypoxic conditions in the presence or absence of TRAP. Data arerepresentative of another experiment. (c) HCT8/S11 cells expres-sing DN-RhoA, DN-Rac1, and DN-Cdc42 were assayed for theirinvasive capacity in collagen type I under hypoxia, in the presenceor absence (Cont: no addition) of TRAP (10mM)

Figure 7 Impact of hypoxic conditions on TN-C and SPARCexpression. Identification of the TN-C and SPARC gene transcripts(a) and proteins (b) in HCT8/S11 cultured for 24 h at 371C on topof collagen type I gels under normoxic or hypoxic conditions. RT–PCR reactions amplified the expected 503 and 217 bp productscorresponding to the TN-C and SPARC amplimers, respectively.GAPDH cDNA was amplified as a loading control (574 bp).Western blot analysis of TN-C and SPARC in HCT8/S11 cellsusing 100mg protein in each lane and 12% SDS–PAGE.Immunodetection was performed using the anti-TN-C mAb BC8(v/v dilution of the hybridoma supernatant) and the anti-SPARCpAb (1 : 1000). Membranes were stripped and reprobed with theanti-HIF-1a mAb (1 : 250) as control for hypoxia. Two majorbands of HIF-1a were detected in normoxic conditions and wereremarkably induced by 24 h hypoxia. These slower and fastermigrated bands are phosphorylated and dephosphorylated formsof HIF-1a, respectively. Data are representative of two otherseparate experiments


8246

Oncogene

We found that the PAR-1 proinvasive activitypromoted by the sGC activator BAY41 is Racdependent and Cdc42 independent (Figure 2c), whereasan inverse situation was observed for the cGMP analog8Br-cGMP (Rac independent/Cdc42 dependent, notshown). This difference can be attributed to thepossibility that sGC and PKG are not the onlyBAY41 and cGMP signaling targets. Alternatively,cGMP levels and PKG enzymes type I (a/b isoforms)and type II might be differentially induced, activated,

and phosphorylated by these two cGMP-inducingagents in terms of duration, extent, and cellularlocalization, as suggested by the differential impact ofthe PKG inhibitor KT5823 in preincubation andsimultaneous addition experiments (Figure 8a–c).

Thrombin activation and PAR-1 expression in peritonealand pleural effusions and their derived cancer cellsin culture

Activation of the coagulation system is an invariableproblem for patients with colorectal carcinoma andother solid tumors, as previously described by Pr.Armand Trousseau. As fibrinolysis is mediated byplasmin, which degrades fibrin clots into D-dimers andfibrin degradation products, the detection of D-dimers isfrequently used to verify the activation of fibrinolysis.Since thrombin cannot be measured directly, prothrom-bin fragment 1þ 2, which is released from prothrombinduring thrombin activation, can be used as an indirectmarker for thrombin activation. We observed thatthrombin is already activated in the serum of themajority of cancer patients, as shown by the detectionof higher prothrombin 1þ 2 levels (1.7071.01 nmol/l),as compared to 0.471.1 nmol/l in serum from normalvolunteers. Interestingly, prothrombin 1þ 2 levels wereremarkably increased to 9.8075.2 nmol/l (Po0.0001) inmalignant effusions of the same patients. Interestingly,malignant effusions from gastric, colorectal, and eso-phageal cancer patients (respectively, samples 523–539–591; dilution range of the initial fluid, 1 : 9–1 : 15)induced the invasive response in HCT8/S11 cells(invasion index range: 12–20%, n¼ 3 experiments).As shown in Table 1, PAR-1 expression was positive

by RT–PCR and FACS analysis in cultured cancer cellsisolated from 15 out of 22 malignant effusions (68%)from patients with gastrointestinal or lung cancers,melanoma, or lymphoma. In four peritoneal effusionsfrom colorectal cancer patients, cultured cancer cellswere 100% positive by RT–PCR analysis and 75%positive by FACS analysis.

Discussion

It is now well recognized that cancer cells and the tumorstroma form a ‘new organ’ that promotes invasivetumor growth and metastasis through complex interac-tions between cancer cells with cellular and matricellularcomponents of the tumor, as well as several othereffectors linked to immune and blood cells, hypoxia, andthe angiogenic switch (Porte et al., 1995; Bissell andRadisky, 2001; Liotta and Kohn, 2001). The matricel-lular proteins TN-C and SPARC/BM-40/osteonectinare induced during wound healing, angiogenesis, andinflammatory and neoplastic diseases. Tumor matri-cellular components such as TN-C (De Wever et al.,2004) and SPARC in the present study suppress RhoAactivity, cellular spreading, and focal adhesions, andinduce PAR-1-mediated invasive migration of cancercells. Consistently, high levels of TN-C and SPARC in

Figure 8 Impact of cGMP and PKG inhibition on PAR-1proinvasive activity in hypoxia and normoxia. (a) HCT8/S11 cellsin normoxia were preincubated for 15min in the presence orabsence of the PKG inhibitor KT5823 (2 and 20 mM), and thenexposed to hypoxia and TRAP (10 mM), or control vehicle. (b)

HCT8/S11 cells in normoxia were treated for 24 h with the cGCactivator BAY41 (1mM) and TRAP (10 mM), either alone or incombination. (c) HCT8/S11 cells in normoxia were treated for 24 hwith either TRAP (10mM), PAF (0.1mM), or the Met activatorHGF (10U/ml), alone or combined with 8Br-cGMP (5mM).Collagen type I gel invasion was then measured using our standardassay in these respective conditions, in comparison with controlexperiments performed without additions in normoxic and hypoxicconditions (Control: Cont)


8247

Oncogene

many solid tumors correlate with the higher incidence ofmetastases and a poor prognosis (Porte et al., 1995;Ledda et al., 1997; Thomas et al., 2000; Maschler et al.,2004). We also identified a complex signaling interplaybetween the cGMP/PKG cascade, the RhoA–ROK axis,and hypoxia in the determination of the proinvasiveactivity of PAR-1 through heterotrimeric G-proteinsignaling. In support of this model, recent studiesdemonstrated that NO and cGMP pathways inactivatethe RhoA–ROK axis (Sandu et al., 2001) involved inRac1 inhibition, suggesting that upstream elementsusing this cascade function as potential Rac1 activators.The proinvasive small GTPase Rac is activated by PKG(Hou et al., 2004) and G-protein coupled receptors(GPCR) using Gaq subunits, such as PAR-1 (Boodenet al., 2002). Small GTPase RhoA is involved in cellmigration via its direct effector ROK, which promotesMLC phosphorylation, through MLC phosphataseinhibition and direct MLC phosphorylation (Kaibuchiet al., 1999). MLC phosphorylation is also induced byPAR-1, and the MAPK- and calcium-activated MLCKcascade (Hansen et al., 2000; Nguyen et al., 2002).

Reducing RhoA activity has two opposing effects: itpromotes migration by lowering adhesion and decreasescell migration by inhibiting cell body contraction.Consistently, the DN-ROK interfering mutant in thepresent study abrogated several forms of invasioninduced by pS2, leptin, and insulin. This picture furtherillustrates our model of the reciprocal RhoA–Rac1crosstalk and antagonism as a molecular sensor forPAR-1 commutation in cancer cell invasion. Integrationof these counteractive mechanisms can originate fromactin-associated molecular scaffolds containing guaninenucleotide-exchange factors (GEF), as shown for Trioand juxtaposition of GEF for RhoA/Rac1 smallGTPases at cell migration (Bellanger et al., 2000).We have observed that calcium translocations in-

duced by TRAP are preserved in HCT8/S11 cellsexposed to the Rho–ROK inhibitors C3T, BAY41,and Y27632 (not shown), suggesting that these PAR-1commutators are acting downstream the formation andrelease of functional heterotrimeric G-proteins complexat activated PAR-1 and Ga subunits. In this area, Rho-GTPases play pleiotropic roles in cell movements,polarity, vesicle trafficking, and EGF-R processingand degradation, as recently demonstrated for Cdc42(Cerione, 2004). GPCR controlled by thrombin, bom-besin, neurotensin, endothelin, and LPA are alsoimplicated in metalloprotease-mediated EGF-R trans-activation (Darmoul et al., 2004; Schafer et al., 2004;Zhao et al., 2004), a major signaling platform in invasivegrowth (Rodrigues et al., 2003). More recently,the fibroblast-derived matrix metalloprotease MMP-1in the stromal-tumor microenvironment has beendesigned as a new PAR-1 activator that promotesinvasion and tumorigenesis of breast cancer cells (Boireet al., 2005). Moreover, MMP-1 is a new breast cancerpredictive marker to identify patients with lesions thatmay develop into cancer (Poola et al., 2005). There aresome early clues suggesting that the RhoA–ROK axis isalso connected with JNK- and c-Jun/AP-1-dependenttranscription (Marinissen et al., 2004) involved inseveral proinvasive pathways, including src and Wnt(Rivat et al., 2003; Le Floch et al., 2005). In the presentstudy, we have clearly shown that several PAR-1commutators are acting as invasion promoters throughthe cGMP/PKG cascade and inhibition of the RhoA/ROK axis. This finding is reminiscent of the ability ofmacrocyclic alkaloid analogs to exert tumor invasioninhibition through persistent Rho activation (McHardyet al., 2004). Further studies on PAR-1 signaling targetsare, therefore, needed to delineate the relative contribu-tion of these integrative levels on invasion andmetastasis by real-time kinetics of the activation statusof signaling elements involved in cellular invasion andsurvival, including proteomic screens and gene expres-sion profiling.Understanding the pathogenesis of cancer invasion

and metastasis at cellular and molecular levels consti-tutes a major challenge for the development of efficienttherapeutic strategies against neoplastic progression,recurrence, and death. We have observed that PAR-1are expressed in metastatic cancer cells isolated from

Table 1 Expression of PAR-1 by RT–PCR and FACS analysesin cultured cancer cells isolated from peritoneal, pleural, andpericardial effusions from patients bearing colorectal tumors or other

malignancies

Diagnosis Effusion sites PAR-1 expression Patientno.

RT–PCR FACS

Gastrointestinal cancersColorectal carcinoma Peritoneum + + 539Colorectal carcinoma Peritoneum + � 546Colorectal carcinoma Peritoneum + + 534Colorectal carcinoma Peritoneum + + 568Gastric carcinoma Peritoneum + + 523Pancreatic carcinoma Peritoneum + + 496Pancreatic carcinoma Peritoneum + + 467

Lung and pleural cancersLung cancer (NSCLC) Pleura � � 550Lung cancer (NSCLC) Pleura � + 499Lung cancer (NSCLC) Pleura + + 524Lung cancer (NSCLC) Pleura + + 459Lung cancer (NSCLC) Pleura + + 535Lung cancer (NSCLC) Pleura + + 552Lung cancer (SCLC) Pleura + � 551Lung cancer (SCLC) Pleura + + 506Lung cancer (SCLC) Pericardium � + 556Pleural mesothelioma Pleura + � 548Pleural mesothelioma Pleura + + 527Pleural mesothelioma Pleura + + 558

Miscellaneous cancersMammary carcinoma Peritoneum + + 545Melanoma Pleura � � 536Non-Hodgkin’s lym-phoma

Pleura + + 533

PAR-1 mRNA and cell surface expression was determined by RT–PCR and FACS analysis. PAR-1 RT–PCR has been considered aspositive when amplicons are clearly detected after 30 amplificationcycles. At least 10 000 cells were examined by FACS analysis usinganti-PAR-1 mouse mAb (clone WEDE15, Immunotec/Coulter,Marseille, France). + and � indicate positive and negative expressionsignals


8248

Oncogene

peritoneal and pleural effusions in patients with coloncancer and other carcinomas. In the majority of cancerpatients, we found that coagulation and thrombinactivation markers such as d-dimers or prothrombinfragment 1þ 2 are remarkably elevated in serum.Moreover, up to 30% of patients developed manifestdeep vein thrombosis during the natural course of theirdisease, according to the interconnection between cancerand coagulation disorders (Xie et al., 2005) referred toas the Trousseau disorder in 1865. Accordingly, wefound high levels of activated thrombin as well as fibrin,a product of fibrinogen degradation by thrombin, inpleural and peritoneal effusions from colorectal cancerpatients and other malignancies. Overall, we proposethat thrombin and PAR-1 might participate, at least inpart, in metastasis of human solid tumors in perito-neal and pleural cavities. In agreement, HCT8/S11cell invasiveness was induced by pleural and peritonealeffusions from cancer patients with digestive tumors.Thus, the thrombin/PAR-1 connection in the presentstudy may play additional roles with the HGF/Metconnection to cyclooxygenase and prostanoid formation(Boccaccio et al., 2005) in cancer-associated metastasisand thrombosis. It is anticipated that the design of newtherapeutic strategies targeting invasive tumor growthshould be defined in view of the diversity andredundancy of the proinvasive networks connected withPAR-1 signaling in the tumor stroma and regarding thematricellular and oncogenic context of a given tumor incancer patients.

Materials and methods

Cell culture, expression vectors, and stable transfections

The human colon cancer cells HCT8/S11 were maintained inRPMI 1640 (Gibco BRL, Cergy Pontoise, France) supple-mented with 10% fetal calf serum (FBS, PAA LaboratoriesGmbH, Pasching, Austria) plus L-glutamine and antibiotics(Gibco BRL). Human colon cancer cell lines HT29 andHCT116 were cultured in DMEM (Life Technologies Inc.,Cergy Pontoise, France) supplemented with 10% FBS andantibiotics. The cDNAs encoding the myc-tagged dominantactive (CAT, i.e. DC-ROK) or dominant negative ROK (DN-ROK) lacking the PH and Rho-binding domains (DRB/PH)were previously described (Kaibuchi et al., 1999). The vectorspcDNA3 encoding the HA-tagged dominant negative RhoA(T19N, DN-RhoA), dominant negative Rac1 (T17N, DN-Rac1), dominant negative Cdc42 (T17N, DN-Cdc42), andconstitutive active RhoA (G14V, DC-RhoA) were from TheGuthrie cDNA Resource Center (Sayre, PA 18840, USA).HCT8/S11 cells were stably transfected according to describedmethods (Nguyen et al., 2002). MDCKT23 cells expressingmutant small GTPases RhoAV14, RhoAN19, or Rac1V12were a generous gift of Dr WJ Nelson (Jou and Nelson, 1998).

Thrombin activation in peritoneal and pleural effusions fromcolon cancer patients and their derived cancer cells in culture

From November 2002 to November 2004, we collected 58samples of effusions from patients admitted to the OncologyCenter (Kiel University Hospital, Germany). Cancer cells wereharvested from eight peritoneal, 13 pleural, and a pericardial

effusion sites. The mean age was 6379.1 years (range: 53–87years). Cancer cells derived from these malignant effusionswere established in long-term culture (6–26 months), usingRPMI 1640 medium containing 2mM L-glutamine, 10% heat-inactivated FBS, 100U/ml penicillin, 100mg/ml streptomycin,and 10% cell-free original effusion fluid. Repetitive cytopatho-logical examinations included cell morphology by phase-contrast microscopy, determination of mitotic rates andepithelial markers (e.g. human epithelial EpCAM-antigenBerEP4), absence of contaminating lymphocytes and mesothe-lial cells by immunocytochemistry (e.g. leukocyte commonantigen LCA and thrombomodulin, respectively), usingspecific antibodies from Dako (Hamburg, Germany). Quanti-tative determination of prothrombin fragment 1þ 2 in theserum and malignant effusions was performed by ELISA(Behring GmbH, Germany).

Collagen type I invasion assays and hypoxia

Collagen type I invasion by cancer cells was monitored aspreviously described (Bracke et al., 2000). The invasion indexis the percentage of cells invading the collagen gel overthe total number of cells. None of the compounds tested in the24 h assay interfered with cell growth and viability. Whereindicated, HCT8/S11 cell cultures were placed for 24 h at 371Cin a sealed incubator chamber device (OXOID Ltd, Basing-toke, Hampshire, England) containing an AnaeroGen baglowering oxygen levels below 1% within 30min.

Western blot and immunofluorescence analyses

For immunoblotting, cultured cells were homogenized at 41Cin lysis buffer containing phenylmethylsulfonyl fluoride(PMSF), dithiothreitol (DTT), aprotinin, pepstatin A, andleupeptin as protease inhibitors, as previously described(Nguyen et al., 2002). The blots were probed with one of thefollowing antibodies: anti-HA monoclonal (mAb) clone12CA5 (1 : 100; Roche Laboratories, Meylan, France), anti-myc mAb 9E10 (1 : 500; Santa Cruz Biotechnology, CA, USA),mAb anti-thrombin receptor (1 : 4000; clone IIaR-A) labelingthe 90 kDa PAR-1 receptor (Biodesign International, Kenne-bunk, ME, USA), anti-TN-C supernatant clone BC8 (v/v)kindly provided by Dr L Zardi (Laboratory of Cell Biology,Genoa, Italy), anti-SPARC goat antibody (1 : 1000; R&DSystems Europe Ltd, Lille, France), and anti-HIF-1a mAb(1 : 250; Becton Dickinson France SA, Le Pont de Claix,France). Membranes were probed for 1 h with peroxidase-linked goat anti-mouse IgGs (1 : 2000; Santa Cruz Biotechno-logy, CA, USA) and revealed by enhanced chemiluminescence(ECL plus, Amersham Pharmacia Biotech). For F-actin andvinculin staining, HCT8/S11 cells were seeded under serum-free conditions on coverslips coated with collagen type I(10 mg/ml) or SPARC (10�8M). Cells were then incubated for1 h with phalloidin-FITC or the anti-vinculin mAb clonehVIN-1 (1 : 400; Sigma), and incubated for 1 h with asecondary antibody (Alexa Fluor 488 anti-mouse, 1 : 400;Molecular Probes, Eugene, USA). Samples were analysed witha Leica DMR fluorescence microscope (Leica MicrosystemesSA, Reuil-Malmaison, France) and a Nikon Eclipse TE300inverted phase-contrast microscope (Micromecanique, Evry,France). For the detection of cell surface PAR-1 in malignanteffusions-derived cancer cells, a FACS scan analysis wasperformed. About 5� 105 cells were washed twice in ice-coldPBS containing 0.05% sodium azide in PBS, and incubated for1 h at 371C with anti-PAR-1 mouse mAb, clone WEDE15(2 mg, i.e. 1 : 100; Immunotec Coulter, Marseille, France).Cells were incubated for 1 h at 371C with anti-mouse


8249

Oncogene

FITC-conjugated secondary antibody (1 : 1000, Dako), andresuspended in 1ml PBS for FACS analysis. Cells incubatedwithout primary antibody followed by labeling with secondaryantibody were used as controls. Flow cytometric analysis wascarried out using a ‘Galaxy Argon Plus’ and the results wereanalysed with ‘Flomax Software’ (Dako). At least 10 000 cellswere examined for each determination, and expression o10%has been considered as negative.

GTP-bound RhoA and Rac1 pull-down assays

HCT8/S11 cells were washed twice with ice-cold PBS and lysedin Mg2þ lysis/wash buffer (25mM Hepes, pH 7.5, 150mMNaCl, 1% Igepal CA-630, 10mM MgCl2, 1mM EDTA, 2%glycerol, and 10mg/ml leupeptin or aprotinin). Cell lysateswere cleared by centrifugation (14 000 g , 5min, 41C), andequal volumes of lysates were incubated for 45min at 41C witheither Rhotekin RBD-agarose beads (Upstate Biotechnology,Campro Scientific) or PAK-1-agarose beads (25 mg). The beadswere washed four times with Mg2þ lysis/wash buffer. GTP-bound and total levels of RhoA and Rac1 proteins weredetected by immunoblotting (15% SDS–PAGE), using a mAbagainst RhoA (Santa Cruz Biotechnology) or Rac1 (Transduc-tion Laboratories, Lexington, KY, USA). The relativeintensity was determined with the ImageQuant software(Amersham Biosciences).

cGMP radioimmunoassay and intracellular free Ca2þ

concentration

HCT8/S11 cells were seeded in 35mm Petri dishes (200 000cells per dish) and incubated overnight with the indicatedeffectors. cGMP production was measured by radioimmu-noassay (Amersham Biosciences). Changes in HCT8/S11intracellular Ca2þ concentrations in response to TRAP weremonitored using the QuantiCell 700 dynamic imaging micro-scopy system (Visitech International Ltd, UK), as described(Ricort et al., 2002). The cells were cultured overnight on glasscoverslips in a six-well plate (200 000 cells per well), in thepresence or absence of C3T, BAY41, or Y27632. Then, controland treated cells were washed twice and incubated for 2 h at371C in PBS-calcium-free HEPES medium containing 5mMFURA2/AM (Molecular Probes Inc., Eugene, USA). Afterbackground recording for 40 s (20 images), the experiment wasinitiated by adding TRAP (20 mM).

RNA isolation and RT–PCR amplification

Total RNA was extracted using the Trizol reagent (Invitrogen)or the ‘Qiagen RNeasy Mini Kit’ (Qiagen). RT–PCR analyseswere performed with the SuperScript One-Step RT–PCR kit(Invitrogen). Fragments of the human TN-C and SPARCcDNAs were amplified using the following primers: TN-C: 50-CCC TGC AGT GAG GAG CAC GGC ACA-30 and 50-TGCCCA TTG ACA CAG CGG CCA TGG-30; SPARC: 50-AAGATC CAT GAG AAT GAG AAG-30 and 50-AAA AGCGGG TGG TGC AAT-30. The amplified amplicons (503 and217 bp, respectively) were resolved in 1.5% agarose gel stainedwith ethidium bromide. For PAR-1 gene expression, thefollowing PCR primers and conditions were used: F, 50-GTGCTGTTTGTGTCTGTGCT-30; R, 50-CCTCTGTGGTGGAAGTGTGA-30 (30 cycles, annealing temperature 551C, 598 bpproduct); nested F, 50-GGGCTTCCTTCACTTGTCT-30; R,50-ACTTCTTGCTGCGGTTGG-30 (30 cycles, 541C, 273 bp);and b-actin F, 50-ATCTGGCACCACACCTTCTACAATGAGCTGCG30, R, 50CGTCATACTCCTGCTTGCTGATC

CACATCTGC-30 (16 cycles, 581C, 838 bp product). Amplifi-cations were performed using a Perkin Elmer Cycler (AppliedBiosystem 2400) and 30ml reaction mixtures containing 1.5mMMgCl2, 0.2mM deoxynucleotide triphosphate, 0.5 mM eachprimer, and 1.5U of Taq DNA polymerase (Invitrogen). Afteran initial denaturation step (5min at 941C), thermal cyclingwas performed for 30 cycles (1min denaturation at 941C, 1minannealing at 551C, 2min synthesis at 721C), followed by a7min final extension step. A 1ml portion of the first-roundPCR product was reamplified in the nested PCR at thefollowing profiles: 941C for 30 s, 541C for 30 s, 721C for 1minand 30 cycles. PCR products were electrophoresed on agarosegels, stained with 5mg/ml ethidium bromide, and visualizedunder UV light.

Peptides and chemicals

The thrombin PAR-1 receptor activating peptide TRAP(hexapeptide SFLLRN), PAR-2 and PAR-4 agonists(SLIGRL and GYPGQV, respectively), platelet activatingfactor (PAF), and insulin were purchased from BachemBiochimie (Voisins-le-Bretonneux, France), Neosystem (Stras-bourg, France), and Sigma-Aldrich (Saint-Quentin Fallavier,France), respectively. The direct activator of sGC BAY41-2272 (abbreviated as BAY41 in the present study), the ROKinhibitor Y27632, and the Clostridium botulinum exoenzymeC3 transferase (abbreviated as C3T), which ADP-ribosylatesand inactivates the small GTPases RhoA, B, and C, weregenerous gifts from Dr J-P Stasch (Wuppertal, Germany),Yoshitomi Pharmaceutical Industries Ltd (Osaka, Japan), andDr Gilles Flatau (INSERM U627, Nice, France), respectively.TN-C and SPARC were from Chemicon International Inc.(Temecula, CA, USA) and Calbiochem, respectively. Inhibi-tors targeting CaM-MLCK, PLCb, and proteasome (KT5926,U73122, and MG132) were from Calbiochem (Meudon,France); 8-bromo-cGMP and the PKG inhibitor KT5823were from BIOMOL Res. Labs. (Tebu- Bio, France); leptinwas from R&D Systems Europe Ltd; and recombinant hpS2(TFF1) produced in Escherichia coli was generously providedby Dr B Westley (University of Newcastle upon Tyne, UK).

Statistical analyses

Significance between experimental values was assessed by theunpaired Student’s t-test at P-values o0.05. Data are mean-s7s.e.m. from three to four experiments.

AbbreviationsECM, extracellular matrix; HIF-1a, hypoxia-inducible factor1a; PAR-1, proteinase-activated receptors type I; PKG,protein kinase G; ROK, Rho-associated coiled-coil-containingprotein kinase; TN-C, tenascin-C; SPARC, secreted proteinacidic rich in cysteine.

Acknowledgements

This work was aided by INSERM, IPSEN (PhD grant to Q-DN), and the FORTIS Verzekeringen (Brussels, Belgium). Wethank Dr Rudolph for the characterization of malignanteffusions-derived cancer cells, Miss C Boissard and Mr MClark for valuable technical assistance, and Dr K Kaibuchi,Dr L Zardi, Dr G Flatau, Dr WJ Nelson, Dr J-P Stasch, DrB Westley, and Yoshitomi Pharmaceutical Industries Ltd(Osaka, Japan) for sharing materials.


8250

Oncogene

References

Ahn B and Ohshima H. (2001). Cancer Res., 61, 8357–8360.Amano M, Chihara K, Kimura K, Fukata Y, Nakamura N,Matsuura Y and Kaibuchi K. (1997). Science, 275, 1308–1311.

Andre T, Kotelevets L, Vaillant JC, Coudray AM, Weber L,Prevot S, Parc R, Gespach C and Chastre E. (2000). Int. J.Cancer, 86, 174–181.

Begum N, Sandu OA, Ito M, Lohmann SM and Smolenski A.(2002). J. Biol. Chem., 277, 6214–6222.

Bellanger JM, Astier C, Sardet C, Ohta Y, Stossel TP andDebant A. (2000). Nat. Cell Biol., 2, 888–892.

Bissell MJ and Radisky D. (2001). Nat. Rev. Cancer, 1, 46–54.Boccaccio C, Sabatino G, Medico E, Girolami F, Follenzi A,Reato G, Sottile A, Naldini L and Comoglio PM. (2005).Nature, 434, 396–400.

Boire A, Covic L, Agarwal A, Jacques S, Sherifi S andKuliopulos A. (2005). Cell, 120, 303–313.

Booden MA, Siderovski DP and Der CJ. (2002). Mol. Cell.Biol., 22, 4053–4061.

Bracke M, Boterberg T, Bruyneel E and Mareel M. (2000).Cancer Invasion and Metastasis, Vol. 58, Humana Press:Totowa, NJ, USA, pp. 81–89.

BradshawAD and Sage EH. (2001). J. Clin. Invest., 107, 1049–1054.Cerione RA. (2004). Trends Cell Biol., 14, 127–132.Coughlin SR. (2000). Nature, 407, 258–264.Darmoul D, Gratio V, Devaud H, Lehy T and Laburthe M.(2003). Am. J. Pathol., 162, 1503–1513.

Darmoul D, Gratio V, Devaud H, Peiretti F and Laburthe M.(2004). Mol. Cancer Res., 2, 514–522.

DeWever O, Nguyen QD, Van Hoorde L, Bracke M, Bruyneel E,Gespach C and Mareel M. (2004). FASEB J., 18, 1016–1018.

Emami S and Perry MC. (1984). Biochim. Biophys. Acta, 804,77–88.

Even-Ram S, Uziely B, Cohen P, Grisaru-Granovsky S, MaozM, Ginzburg Y, Reich R, Vlodavsky I and Bar-Shavit R.(1998). Nat. Med., 4, 909–914.

Even-Ram SC, Maoz M, Pokroy E, Reich R, Katz BZ,Gutwein P, Altevogt P and Bar-Shavit R. (2001). J. Biol.Chem., 276, 10952–10962.

Faivre S, Regnauld K, Bruyneel E, Nguyen QD, Mareel M,Emami S and Gespach C. (2001).Mol. Pharmacol., 60, 363–372.

Hansen SH, Zegers MM, Woodrow M, Rodriguez-Viciana P,Chardin P, Mostov KE and McMahon M. (2000).Mol. Cell.Biol., 20, 9364–9375.

Hou Y, Ye RD and Browning DD. (2004). Cell Signal., 16,

1061–1069.Jou TS and Nelson WJ. (1998). J. Cell Biol., 142, 85–100.Kaibuchi K, Kuroda S and Amano M. (1999). Annu. Rev.

Biochem., 68, 459–486.Kamath L, Meydani A, Foss F and Kuliopulos A. (2001).

Cancer Res., 61, 5933–5940.Kawasaki K, Smith Jr RS, Hsieh CM, Sun J, Chao J and LiaoJK. (2003). Mol. Cell. Biol., 23, 5726–5737.

Kisley LR, Barrett BS, Bauer AK, Dwyer-Nield LD, BarthelB, Meyer AM, Thompson DC and Malkinson AM. (2002).Cancer Res., 62, 6850–6856.

Kotelevets L, Noe V, Bruyneel E, Myssiakine E, Chastre E,Mareel M and Gespach C. (1998). J. Biol. Chem., 273,

14138–14145.Kureishi Y, Kobayashi S, Amano M, Kimura K, Kanaide H,Nakano T, Kaibuchi K and Ito M. (1997). J. Biol. Chem.,272, 12257–12260.

Le Floch N, Rivat C, De Wever O, Bruyneel E,Mareel M, Dale T and Gespach C. (2005). FASEB J., 19,

144–147 .

Ledda MF, Adris S, Bravo AI, Kairiyama C, Bover L,Chernajovsky Y, Mordoh J and Podhajcer OL. (1997). Nat.Med., 3, 171–176.

Liotta LA and Kohn EC. (2001). Nature, 411, 375–379.Marinissen MJ, Chiariello M, Tanos T, Bernard O, NarumiyaS and Gutkind JS. (2004). Mol. Cell, 14, 29–41.

Maschler S, Grunert S, Danielopol A, Beug H and Wirl G.(2004). Oncogene, 23, 3622–3633.

McHardy LM, Sinotte R, Troussard A, Sheldon C,Church J, Williams DE, Andersen RJ, Dedhar S,Roberge M and Roskelley CD. (2004). Cancer Res., 64,

1468–1474.Murphy-Ullrich JE, Lane TF, Pallero MA and Sage EH.(1995). J. Cell. Biochem., 57, 341–350.

Nguyen QD, Faivre S, Bruyneel E, Rivat C, Seto M, Endo T,Mareel M, Emami S and Gespach C. (2002). FASEB J., 16,

565–576.Pennacchietti S, Michieli P, Galluzzo M, Mazzone M,Giordano S and Comoglio PM. (2003). Cancer Cell, 3,

347–361.Poola I, DeWitty RL, Marshalleck JJ, Bhatnagar R, AbrahamJ and Leffall LD. (2005). Nat. Med., 11, 481–483.

Porte H, Chastre E, Prevot S, Nordlinger B, Empereur S,Basset P, Chambon P and Gespach C. (1995). Int. J. Cancer,64, 70–75.

Ricort JM, Lombet A, Lassarre C and Binoux M. (2002).FEBS Lett., 527, 293–297.

Ridley AJ. (2001). J. Cell Sci., 114, 2713–2722.Rivat C, Le Floch N, Sabbah M, Teyrol I, Redeuilh G,Bruyneel E, Mareel M, Matrisian LM, CrawfordHC, Gespach C and Attoub S. (2003). FASEB J., 17,

1721–1723.Rodrigues S, Attoub S, Nguyen QD, Bruyneel E, RodrigueCM, Westley BR, May FE, Thim L, Mareel M, Emami Sand Gespach C. (2003). Oncogene, 22, 4488–4497.

Rodrigues S, Nguyen QD, Faivre S, Bruyneel E, Thim L,Westley B, May F, Flatau G, Mareel M, Gespach C andEmami S. (2001). FASEB J., 15, 1517–1528.

Sandu OA, Ito M and Begum N. (2001). J. Appl. Physiol., 91,1475–1482.

Sauzeau V, Le Jeune H, Cario-Toumaniantz C, Smolenski A,Lohmann SM, Bertoglio J, Chardin P, Pacaud P andLoirand G. (2000). J. Biol. Chem., 275, 21722–21729.

Sawada N, Itoh H, Yamashita J, Doi K, Inoue M, MasatsuguK, Fukunaga Y, Sakaguchi S, Sone M, Yamahara K,Yurugi T and Nakao K. (2001). Biochem. Biophys. Res.Commun., 280, 798–805.

Schafer B, Gschwind A and Ullrich A. (2004). Oncogene, 23,

991–999.Stasch JP, Becker EM, Alonso-Alija C, Apeler H, DembowskyK, Feurer A, Gerzer R, Minuth T, Perzborn E, Pleiss U,Schroder H, Schroeder W, Stahl E, Steinke W, Straub A andSchramm M. (2001). Nature, 410, 212–215.

Thomas R, True LD, Bassuk JA, Lange PH and Vessella RL.(2000). Clin. Cancer Res., 6, 1140–1149.

Van Obberghen-Schilling E, Vouret-Craviari V, Chen YH,Grall D, Chambard JC and Pouyssegur J. (1995). Ann. NYAcad. Sci., 766, 431–441.

Wojtukiewicz MZ, Tang DG, Ben-Josef E, Renaud C, WalzDA and Honn KV. (1995). Cancer Res., 55, 698–704.

Xie WZ, Leibl M, Clark MR, Dohrmann P, Kunze T andGieseler F. (2005). Biomed. Pharmacother., 59, 70–75.

Zhao D, Zhan Y, Koon HW, Zeng H, Keates S, Moyer MPand Pothoulakis C. (2004). J. Biol. Chem., 279, 43547–43554.


8251

Oncogene

Documents

Commutators of PAR-1 signaling in cancer cell invasion reveal an essential role of the Rho–Rho kinase axis and tumor microenvironment